Research Paper No. 2006/104 Credit Constraints as a Barrier to Technology Adoption by the Poor Lessons from South-Indian Small-Scale Fishery Xavier Giné1 and Stefan Klonner2 September 2006
Abstract
We study the diffusion of a capital intensive technology among a fishing community in south India and analyze the dynamics of income inequality during this process. We find that lack of asset wealth is an important predictor of delayed technology adoption. During the diffusion process, inequality follows Kuznets’ well-known inverted U-shaped curve. The empirical results imply that redistributive policies favouring the poor result in accelerated economic growth and a shorter duration of sharpened inequality.
Keywords: technology adoption, inequality, fishing sector, India
JEL classification: O33, O13, O25
The World Institute for Development Economics Research (WIDER) was established by the United Nations University (UNU) as its first research and training centre and started work in Helsinki, Finland in 1985. The Institute undertakes applied research and policy analysis on structural changes affecting the developing and transitional economies, provides a forum for the advocacy of policies leading to robust, equitable and environmentally sustainable growth, and promotes capacity strengthening and training in the field of economic and social policy making. Work is carried out by staff researchers and visiting scholars in Helsinki and through networks of collaborating scholars and institutions around the world.
UNU World Institute for Development Economics Research (UNU-WIDER) Katajanokanlaituri 6 B, 00160 Helsinki, Finland Camera-ready typescript prepared by the authors. The views expressed in this publication are those of the author(s). Publication does not imply endorsement by the Institute or the United Nations University, nor by the programme/project sponsors, of any of the views expressed.
The Tables and Figures appear at the end of the paper.
1 Introduction
Globalization has a¤ected the livelihoods of �shing communities in south Asia in several
ways over the past half century. In this paper we study one facet of these developments,
the adoption of beach-landing �bre reinforced plastic boats (FRP) by �shing households
in Tamil Nadu, India. The di¤usion of this new technology, which replaces traditional
artisan wooden boats, is as much a product of ongoing globalizing trends as it is a response
to distortions caused by previous waves of innovation triggered by globalization.
We shed light on this process by studying both the determinants of technology adop-
tion as well as the resulting income and inequality dynamics over the process of technology
di¤usion within a �shing village. The data, which was collected by the authors in 2002 and
2004, cover 65 boat-owning households of a �shing village where the �rst �bre boats ap-
peared in 2001. We �nd, �rst, that poorer households adopt later while ability to operate
the new technology does not signi�cantly predict the timing of adoption. Thus inequality
and lack of wealth is responsible for a socially ine¢ cient sequence of individual adoptions,
whereby the rich and not the most able �shermen adopt �rst. Qualitative interviews with
respondents suggest that lack of wealth delays technology adoption mainly through credit
constraints and, to a lesser extent, higher risk aversion among poorer households.
Second, we �nd that inequality during the process of technology di¤usion follows
Kuznets�well-known inverted U. Initially, the technological innovation widens the gap
between the rich and the poor, but after the entire community has completed the tech-
nological shift, inequality drops to a lower level than before, which implies that in the
long run the innovation studied here bene�ts the poor more than proportionally. We con-
duct simulations to investigate how di¤erent counterfactual distributions of initial wealth
across the sample a¤ect adoption timings. Here we �nd that a redistributive policy favor-
ing the poor results in accelerated economic growth and a shorter duration of sharpened
inequality, albeit the quantitative impact of such a policy is small. When we simulate the
adoption process for a sample of only rich households, in contrast, the process of adoption
is completed ten times as fast as observed in the actual data, implying that rich commu-
nities can enjoy the bene�t from technological innovation, and thus grow, considerably
1
faster than poor ones. These �ndings provide a micro illustration of Nissanke and Thor-
becke�s (2005) point that the relationship between globalization and poverty is complex
and may be non-linear.
Among existing studies of technology adoption in low income environments, the con-
text studied here is of particular interest because we focus on a capital intensive technol-
ogy. In contrast, the bulk of existing literature has focused on divisible, comparatively
inexpensive technologies, such as high yield variety seeds, the switch from food to cash
crops, or use of chemical fertilizers. As a consequence, the role of wealth and initial in-
equality among a group of entrepreneurs for the adoption process, as well as the resulting
income and inequality dynamics have deserved little attention.
The rest of this paper is organized as follows. In the next section we provide some
background on globalization and India�s �shing sector. Section 3 introduces the context
of this study and the data. Section 4 reviews relevant existing literature on technology
adoption. Section 5 sketches a theoretical framework that illustrates how wealth a¤ects
the timing of technology adoption. Section 6 develops the empirical methodology and
presents results. In Section 7 we simulate the adoption process for alternative distributions
of initial wealth. The �nal section evaluates the �ndings and draws conclusions.
2 Globalization and South India�s Fishing Sector
To put the present study into the more general perspective of globalization and its impact
on the poor, this section sketches important developments in south India�s �sheries over
the last 40 years with particular reference to the consequences of international develop-
ment assistance and technology di¤usion.
Until the 1950s the prevailing vessel on the coasts of southern Kerala and Tamil Nadu
was the kattumaram, a boat which is manufactured by hand tying together a few logs of
wood which are shaped by traditional carpenters. Kattumaram literally means tied-log
raft (maram is Tamil for log while kattu means tied). The timber used for kattumarams is
albizia, a light weight, fast-growing tropical tree found in forests throughout south India.
Traditionally, kattumarams were equipped with a sail for propulsion.
2
South India�s �sheries were hit by globalization as early as the late 1950�s when Eu-
ropean donors implemented large comprehensive development projects. The case of the
Indo-Norwegian project is particularly well documented (e.g. Sandven, 1959), which called
for the mechanization of �shing boats, provision of repair facilities, introduction of new
types of �shing gear, improvement of processing methods, building of ice plants, and sup-
ply of insulated vans and motor crafts for transport of fresh �sh. The most successful
vessel introduced under the program was a fully mechanized 32-ft trawler with a powerful
84-90 hp inboard engine. A new trawler cost around Rs. 125,000 in 1978 prices (which
equals about Rs. 600,000 in 2004) and had a crew of 15-20 members. The high cost of
the gear and limited access to credit explains why the majority of trawler owners were
businessmen and traders rather than genuine �shermen. Owners used to hire a captain
and a crew to operate the vessel and provided incentives by entitling each of them to a
share of the �sh sales.
This and subsequent development projects led to a considerable change in the struc-
ture of asset ownership and labor relations in �shing communities. While family sized
small scale enterprises were the dominant mode previously, productive assets were now
concentrated in the hands of a few. Moreover, economies of scale made much of the
labor previously employed in the �shing sector redundant and many �shermen became
wage laborers on trawlers as the traditional technology was not able to compete with the
new one. In consequence, while aggregate production soared, asset and income inequality
increased as well (Platteau, 1984; Kurien, 1994).
The introduction of mechanized vessels, moreover, has depleted the resource base on
which small as well as large scale �shing was relying by harvesting shrimp in waters
close to the coastline in large quantities. In this connection, it is estimated that Tamil
Nadu currently has as much as twice the number of trawlers that could be sustained by
the resource base on the long run (Vivekanandan, 2002). Since the mid 1980�s, these
developments have increasingly threatened the livelihoods of small-scale �shermen along
the coasts of south India. It should be noted in this connection that many places on
the coast do not have the option to directly engage in trawler �shing because a trawler
requires harbor facilities as, unlike a kattumaram, the vessel is too large to land on a
3
beach.
The depletion of the resource base in waters adjacent to the shore, moreover, increased
the pressure on small-scale �shermen to venture into deeper waters. These developments,
in turn, created rising demand for engine propulsion in the form of an outboard motor
(OBM), which increases the radius of operations of a kattumaram considerably. At this
point, globalization enters the picture once more with India�s federal government easing
the until then heavily protective import policies, which led to a drop in cost of imported,
internationally leading brands, such as Yamaha, Suzuki and Evinrude. It thus comes as
no surprise that since the mid 1980�s small OBMs of eight to nine horse powers have
spread rapidly throughout south India�s coasts.1 It became common practice to mount
such an engine on a kattumaram, which was previously propelled by sail and manpower
only (Kurien, 1994).
Finally, in the mid 1990�s �bre reinforced plastic boats entered the stage. Several
factors simultaneously contributed to this development. First, the technological hybrid
of kattumaram and OBM proved to be problematic as the vibrations of the engine strain
and damage the substance of the vessel (Kurien, 1995). Second, the material used for
FRP production became cheaper relative to the timber used for kattumaram manufac-
turing. On the one hand, through trade liberalization, �bre materials, which had been in
use in the western hemisphere in aerospace, automotive and marine industries since the
1950�s, became less costly. On the other hand, albizia became more and more scarce and
expensive because of successive deforestation and other demands. Finally, blueprints for
appropriate shapes of FRP boats (capable of negotiating high surf and beach landing)
became available. In 1995, a boat yard near Pondicherry started out by manufactur-
ing a boat, the so-called Maruthi boat, which resembles a vessel previously developed in
Sri Lanka for similar coastal conditions as encountered in southern Tamil Nadu (Kurien,
1995). Moreover, supported by federal funds, the Tamil Nadu state government sponsored
research and development of a new model particularly suited for maritime conditions of In-
dia�s south-eastern coasts during the 1990�s, which went into production in 2000 (Pietersz,
1In contrast, previous attempts of leading international OBM manufacturers to target India�s small
scale �shermen in the 1970�s were largely unsuccessful (Pietersz, 1993).
4
1993; The Hindu, 2001). Both of these boat types are around 18 feet long and are operated
by a crew of three to four �shermen.
The combination of FRP boat and OBM facilitates a considerably wider radius of
operation than the kattumaram as well as greater carriage capacity and more convenience
(The Hindu, 2001). It is also worth noting that the emergence of the beach landing FRP
has left the labor-intensive character and fragmented ownership of productive assets of
kattumaram �sheries unchanged. In contrast to the great societal changes triggered by the
earlier development programs, FRPs thus appear to have the potential to improve indi-
vidual livelihoods without turning the distribution of productive assets and the structure
of labor markets upside down.
3 The Study Village
The village of study is located in the southern part of the coast of the gulf of Bengal, close
to the pilgrim center of Tiruchendur. With a population of 1,500, there were 75 boats
operated by 67 households in late 2003. About 250 men worked on these boats, either
as owner/captain, family crew or wage laborer. The village has neither a harbor nor a
jetty, a fact that restricts operations to beach-landing boats. All year-round operating
vessels have a crew of two to four men and are operated by local households. All of these
households belong to the exclusively catholic boat-owning community of the village, which
used to belong to a speci�c caste before collectively converting about 400 years ago.
On a typical day, boats leave the shore around 1 am and land at the village�s market
place on the beach between 7 and 11 in the morning. There, local �sh auctioneers market
the catches to a group of buyers, which comprise local traders as well as agents of nation-
wide operating �sh-processing companies.
In our study village, the �rst FRPs were adopted in January 2001. By January 2004, 48
households were operating at least one FRP. The vast majority of FRPs is of the Maruthi
type and 18 feet long by 7 feet wide, with two boats being slightly longer, measuring
21 � 7 feet. According to villagers, FRPs started to spread in 2001, but not earlier,because an FRP dealership opened in nearby Tiruchendur around that time, making such
5
boats readily available. The cost of a vessel is around Rs. 70,000. All of the adopting
households already owned a seven to nine horse powers OBM previously, which sells at
Rs. 50,000 to 70,000. In comparison, a new kattumaram came at a cost of around Rs.
20,000 at the time of our 2004 interview.
According to �shermen and our data, with the same number of crew, an FRP�s landings
are about 50% bigger than those of a kattumaram. Given the yields of �bre-boat �shing,
every owner of a kattumaram in the village we interviewed assured that he wanted to
switch to a �bre boat as soon as possible. Fishermen repeatedly pointed out, however,
that �shing on an FRP requires a di¤erent set of skills than those needed to operate
a kattumaram. For that reason it is common practice among the buyers of �bre boats
in the village to hire migrant laborer-�shermen from Kerala as crew members who have
previously gathered experience with this technology.
Vessel �nancing and marketing of �sh catches are interlinked for almost all boat owning
households that we interviewed. Although the focus of the present study is on the adoption
of FRPs, it is instructive to start out with the credit cum marketing contract common for
kattumarams. For the purchase of a craft, the auctioneer gives a loan of about Rs. 15,000
and 25,000. In return, the boatowner sells all daily catches through that auctioneer, who
keeps 5 percent of the value of the sales. The boatowner does not repay the principal.
As a consequence, the commission comprises a compensation for the marketing services
as well as an implicit interest payment on the amount owed. When a boatowner switches
auctioneers, the new auctioneer settles the debt with the previous one. Switching of
auctioneers does occur occasionally. The superiority of this interlinked share arrangement
over separate debt and marketing contracts is likely a result of, �rst, limited liability of
the �sherman and, second, costless monitoring of the �sherman�s day-to-day success by
the auctioneer. It is interesting to note that this credit cum marketing arrangement is
identical to the one reported by Platteau (1984) in �shing villages in Kerala twenty years
earlier.
The contract for FRP �nancing is similar, albeit not identical. The auctioneer ad-
vances funds for the purchase of the vessel. However, in addition to a commission of 7
percent, the auctioneer keeps another 10 percent of daily sales, which he deducts from the
6
principal owed by the boatowner. Unlike a kattumaram owner whose level of debt remains
constant, an FRP owner asks his auctioneer for additional funds from time to time. When
such additional funds are granted, they bare no interest and are added to the �sherman�s
outstanding balance. The emergence of this feature of debt reduction and repeated rene-
gotiation can be explained by the following two reasons. First, �bre boat �shing consumes
more working capital, such as nets. To cover these costs, the owner of an FRP has to
incur expenses between Rs. 5,000 and 20,000 from time to time. Second, since the FRP
is a new technology, each individual�s ability to operate it is not precisely known initially.
Since the auctioneer�s cash-�ow directly depends on the �sherman�s day-to-day success,
however, the debt reduction component allows the auctioneer to drive down the debt level
of an ex-post unsuccessful �sherman to a level at which the auctioneer�s opportunity cost
of capital does not exceed his commission income.2 Many �shermen interviewed stated
that the funds extended by the auctioneer initially do not su¢ ce to cover the entire cost of
the technology switch. It was, moreover, stated that bank and even money lender credit
is virtually unavailable for this purpose as these lending sources do not accept a boat as
collateral. Savings were, therefore, mentioned as the second most important source of
funds to cover the cost of a �bre boat.
We brie�y discuss the structure of labor contracts. On kattumarams, in Platteau�s
as well as our study village, typically at least two members (two brothers or father and
son) of the family which owns the vessel sail on the boat. The rest of the crew consists
of laborer-�shermen. To ensure daily availability of non-family labor, boatowners often
tie laborers by advancing interest-free credit. On FRP boats, the common remuneration
scheme for laborers is based on shares. Speci�cally, from the money which the boatowner
receives from the auctioneer (that is net of commission and debt reduction), the expenses
for fuel (around Rs. 200 per day on average) are deducted. The remainder is divided
equally. One half goes to the boatowner. The other half is equally divided among all crew
members who have sailed on the boat that day. If the boatowner sails himself, he also
enjoys one of those shares.
Our data, which we collected between 2002 and 2004, cover all 65 households, which
2This aspect is the subject of a companion paper (Giné and Klonner, 2005).
7
owned and sailed on either a kattumaram or an FRP by the end of 2003.3 We collected
information on the type of vessel operated and the time of adoption of an FRP if ap-
plicable. From auctioneers, we obtained data on monthly �sh sales by household since
2000. We, moreover, conducted a household survey on household demographics and asset
possession. Household level data on �sh sales with a kattumaram were the most di¢ cult
to collect as auctioneers did not always have records dating several years back on �le.
For 26 of the 65 households, however, we were able to collect those data and thus have
a complete picture of sales before and after (if applicable) adoption as well as household
characteristics. This set of households will be referred to as the core sample. Descriptive
statistics for those households are set out in Table 1.
4 Existing Literature on Technology Adoption in Low
Income Countries�Primary Sectors
Much of the literature that studies technology adoption in developing countries concludes
that its pace has been rather slow. Feder et al. (1985), in their excellent review of the
early literature point to factors such as credit constraints, aversion to risk and limited
access to information, to explain why adoption has not been faster. Most of the work they
survey uses static models to explain adoption, while the dynamic properties of adoption
are left to heuristic or comparative-static arguments at best. In particular, the role of
savings, which may be crucial in contexts where credit or insurance markets are imperfect,
especially if the technology is indivisible, does not receive much attention.
The literature distinguishes between divisible technologies, such as high yield varieties
(HYV) or new variable inputs, and indivisible technologies, such as tractors or the one
we study here, FRP boats. If the technology is divisible, one can study the intensity
of adoption of a given farmer as well as the aggregate intensity in a region. When the
technology is indivisible, the decision at the individual level is necessarily a dichotomous
3Two households owned FRPs and hired a crew. In both cases, the household�s head primary occu-
pation is not �shing, for which reason we excluded them from the sample.
8
variable and only the aggregate intensity is still continuous. In the case of technologies
that are not capital intensive, like the adoption of high yield variety (HYV) seeds, lack
of credit is not seen as a major constraint. Instead, most of the more recent literature is
concerned with the interaction between learning about a new technology and its di¤usion.
The �rst of these contributions is Feder and O�Mara (1982), who show that aggregate
adoption at each point in time can follow a sigmoid curve. They consider a scale-neutral
risky innovation with risk-neutral farmers holding prior believes about the mean yield of
the new technology.
Besley and Case (1994) proceed in a similar fashion in their study of the di¤usion of a
new cotton variety in one of the south-Indian ICRISAT villages. In their model, planting
the new variety not only a¤ects current pro�ts but it also generates public information on
the pro�tability of the new versus the old variety. Therefore, there is individual as well
as social learning from planting the new crop. They �nd that adoption occurs with delay
because farmers underestimate initially the technology�s pro�tability and because they
fail to internalize the positive informational externality created by other farmers when
planting the new crop. Among other �ndings, they conclude that wealthier farmers tend
to innovate �rst because the informational externality is largest to them. Poor farmers
adopt later as they bene�t from the positive informational externality generated by rich
farmers.
Foster and Rosenzweig (1995) take for granted that HYV of wheat and rice that
became available during the Indian Green Revolution in the mid 1960�s yield higher
pro�ts than traditional varieties. In their model, however, the pro�tability of HYV�s is
dictated by a target input model, whose optimal level has to be learned. The issue is,
again, individual versus social learning in that each "trial" with the new variety generates
additional information on the optimal level and this information is conveyed not only to
the farmer himself, but also to the entire village (at least to some extent). In contrast to
Besley and Case (1994), however, planting the new crop comes at the cost of choosing an
input level that is far from the optimal, especially in earlier periods when there is little
knowledge about the optimal level. Farmers �nd themselves playing a dynamic public
good game, where each farmer has an incentive to wait because information is generated
9
costlessly by another farmer experimenting with the new crop. As a consequence, those
farmers who expect the greatest bene�ts from experimentation adopt �rst. As in Besley
and Case (1994), those are the relatively wealthy farmers because they operate several
plots, each of which bene�ts from the additional information in future cropping periods.
Interestingly, their results imply that poor farmers in a community of relatively poor
farmers adopt earlier than poor farmers with wealthy neighbors.
Bandiera and Rasul (2004) test for non-monotonicity of information spillovers among
Mozambiquean farmers to whom a new sun�ower variety was made available in 2000.
They �nd an inverted U-shaped relationship between the amount of available information
to a farmer and the probability that he adopts, suggesting that social e¤ects on the
individual adoption decision are positive when there are few adopters in the individual�s
information network, and negative when there are many. Di¤erences in asset wealth are
not found to impact the adoption decision, which is not surprising given that the NGO
that provided the new variety in their context covered all switching costs.
Munshi�s (2003) study of adoption of rice and wheat high yield varieties during the
Indian Green Revolution focuses on the e¤ect of the sensitivity of farm-speci�c growing
conditions on the extent of social learning. He �nds that for rice HYV�s, which are
more sensitive to unobserved farm characteristics than wheat HYV�s, individual adoption
decisions are less responsive to neighbors� experience. His analysis, however, does not
take into account the e¤ect of famers�wealth on their adoption decisions.
To summarize, all of these papers conclude that there is either a positive or no rela-
tionship between individual wealth and the decision to adopt a new technology. Wealth,
however, is typically correlated with, or even indistinguishable from other important indi-
vidual characteristics, such as farm size, education, access to credit, availability of other
inputs, and access to information. Thus, a positive relationship between wealth and early
adoption can be due to alternative factors, which are not disentangled by the existing
empirical analysis. Policy recommendations, however, may well depend on the nature of
the channel through which wealth a¤ects adoption. In the papers focusing on learning, for
example, it is generally argued that poor farmers adopt later because their valuation for
information generated by initial "trials" with the new technology is lower. Thus, an infor-
10
mation campaign about the bene�ts would result in more adoption. In general, however,
it is not clari�ed, whether alternative channels might also play a role. Other potential
candidates are di¤erential risk aversion (see Binswanger et al., 1980), access to capital, or
availability of labor. For example, if the technological innovation is labor intensive and
wealthier households have better access to the labor market, a wealthier household may
adopt earlier just because of labor market conditions. In the present study, we there-
fore make an attempt to thoroughly identify the channel through which wealth a¤ects
adoption decisions.
5 Individual Wealth and Technology Adoption: The-
ory
In this section we sketch a simple model of the propensity to adopt a new, costly technology
and the role of initial wealth in this process. Given the discussion in Section 3, we assume
that agents only have access to a savings technology to accumulate assets. Agents can
produce with a traditional technology (kattumaram) that yields yC or invest in a more
pro�table technology (�bre boat) which yields yF in expectation. The �bre boat can be
purchased at cost K. Since there is no possibility of borrowing, the investment of K must
come from own resources. In line with Section 3 we may think of K as the cost of the
boat net of the loan from the auctioneer and of yt as income net of debt repayment and
commissions. Agents accumulate assets in the following manner,
at+1 = yt � ct + (1 + r)at;
where r is the interest rate on savings, at is the level of assets or liquid wealth in period t,
and ct denotes consumption in period t. We assume that agents start in the �rst period
with an endowment of assets a0.
To keep things simple we assume that agents are risk neutral, live in�nite periods and
discount the future at rate 11+r. Each period, a household decides whether to purchase the
�bre boat and how much to save for the following period. More formally, a household�s
11
task is to choose the vector of next period�s assets fat+1g and the adoption date t� to
where {f�g denotes the indicator function.The program which solves this problem depends on the relative pro�tability of the
new versus the old technology. In particular, if
yF > yC + rK (1)
the optimal program involves saving all income until at � K and switching to the new
technology in that same time period, which gives
t� =ln�rK+yCra0+yC
�ln(1 + r)
;
ct = 0 8t < t�, ct = yF ;8t � t�.When yF � yC + rK, on the other hand, the optimal program involves dissaving
instantly, c0 = yC + a0, and consuming all income generated with the old technology
concurrently, ct = yt for all t > 0.
By di¤erentiating the optimal adoption time t� with respect to the di¤erent parameters
of interest, it is easy to see that the higher the initial level of assets a0, the higher
the income from the kattumaram yC , and the higher the interest rate r, the earlier the
adoption time t�. In this simple setup, t� does not depend on yF other than trough (1).
When utility is concave, however, it can be shown that t� is, moreover, decreasing in yF .
Finally, if several �shermen pool their savings, e.g. through a Rosca, adoption can occur
earlier on average. It continues to hold, nevertheless, that a group of wealthier individuals
can achieve an earlier adoption time on average.
12
6 Estimation
In this section, we seek to empirically identify the determinants of the timing of technology
adoption. As developed in the previous section, a risk-neutral �sherman seeks to adopt the
new technology as quickly as possible when he expects the technology switch to increase
his income. An important explanatory variable for the adoption decision is therefore
the expected change in income resulting from the technology shift. If expectations are
unbiased, the ex-post change in observed income for �sherman i can be interpreted as the
(most likely noisy) realization of i�s expectations. We therefore �rst estimate the income
change of each �sherman who adopted a �bre boat before the interview date and use these
results in the subsequent analysis of the timing of adoption.
6.1 Estimating the Income Change from Adoption
The goal of this section is to provide estimates of the average income that a �shing
household earns with the old and new technology. With the share system that exists
in the village for the compensation of both laborers and the capital obtained from an
auctioneer, household income is roughly proportional to monthly �sh sales generated
by that household. Since both catch quantities as well as daily �sh prices are subject to
substantial �uctuations, however, the following analysis aims at netting out the individual-
speci�c component in how successfully each technology is operated by a given household.
Moreover, we have to allow for the possibility of both individual and social learning when
the new technology is used.
Learning by doing implies that individual catches trend upwards after adoption as
the individual learns how to use the new technology more e¢ ciently over time. Social
learning (or learning from others), on the other hand, implies that an individual can
use the expertise other individuals have acquired with the new technology to become
more e¢ cient himself. Quite generally, the latter implies that the �learning curve�of an
individual, that is his success as a function of time since adoption, depends on the amount
of information available at the time he adopts. More speci�cally, the learning curve of a
later adopter is �atter as he starts out with relatively more information at the time of
13
adoption. With monthly sales data from 43 �shermen who switched to a �bre boat before
the date of the interview, a test for individual as well as social learning is thus facilitated
by the regression speci�cation
log(ysit) = �si + �t + {ft � t�i g� 1�i + 2�
2ii + �1t
�i �i + �2t
�i �2i
�+ usit; (2)
where ysit denotes monthly sales (in Rupees) of �sherman i in month t who currently
operates technology s, where s = C for kattumaram and s = F for a �bre boat. Also
consistent with the notation in the previous section, t�i denotes the time of adoption by
individual i, and �i denotes time since adoption, so that t = t�i + �i. �si is an individual-
speci�c, technology-dependent �xed e¤ect, while �t is a month-speci�c dummy that picks
up aggregate �shing conditions and shocks. Finally, usit is an i.i.d. error term with
E[usit] = 0.
This parametrization assumes that shocks a¤ect sales generated through the old and
new technology identically in a proportional sense. This is strictly true as far as price
�uctuations (per kg of �sh) are concerned as the price indices faced by kattumaram and
�bre boat �shermen are the same. Whether it is also an appropriate assumption for
weather shocks remains an open question. It is to be expected, however, that at least the
sign of the shock works in the same way for both technologies.
While speci�cation (2) does not allow for learning by �shermen who are operating
the old technology, which has been used over several decades, the term 1�i+ 2� 2i allows
for learning by doing for �bre boatowners. In that case, 1 is larger and 2 smaller than
zero if learning by doing exhibits positive and decreasing marginal returns (Foster and
Rosenzweig, 1995). The term �1t�i �i + �2t
�i �2i captures the possibility of learning from
others by allowing for a di¤erent shape of the learning curve for later adopters. Here time
since adoption is interacted with a proxy for the amount of information available at the
time of adoption by individual i, namely the time between the �rst adoption in the village
and the adoption of individual i. With learning from others, the individual learning curve
for a later adopter is �atter as he starts out with more information in hand than any
adopter before him (Foster and Rosenzweig, 1995).
14
A test of the hypothesis of no learning by doing is thus
HL : 1 = 2 = 0:
Analogously, a test of the hypothesis of no social learning is implemented by testing the
composite hypothesis
HS : �1 = �2 = 0:
The results of the estimation of Equation 2 together with F-test statistics for HS and
HL are set out in Table 2. According to these results, the null hypotheses of no social
and no individual learning are rejected, at least at the 10% level. According to the point
estimates of 1 and 2, the �rst adopters in the village experience an increase in sales for
roughly the �rst ten months with the new technology.4 The estimate of �1 on the other
hand implies that the individual learning curve starts out �at for a �sherman who adopts
a �bre boat 12 months after the �rst adoption in the village (the absolute value of b�1equals roughly one twelfth of b 1).We use the insights from the previous estimation for deriving a more restrictive econo-
metric speci�cation, in which there is (positive) individual learning before some cuto¤
date and none of it afterwards. More speci�cally, we estimate
Here t0 denotes the month of the �rst adoption in the village while � is a cuto¤ month
(counted from the time of the �rst adoption in the village), after which no increase in
individual sales occurs. The shape of the Di function can be explained simply: for �sher-
men who adopted no later than � months after the �rst adoption in the village, Di equals
a straight line with slope one before date t�0 + �. From t0 + � onwards, it remains at the
level attained in that period.
4This is obtained by calculating the maximum of the parabola implied by 1 and 2,b 12b 2 .
15
Estimation of (3) by OLS yields a point estimate of � = 5, which implies that learning
by doing occurs during roughly the �rst half year of using the new technology.5 This
is not surprising given that, in contrast to the duration of an agricultural cultivation
cycle, �shing is a daily, and thus a high-frequency activity.6 The full estimation results
for equation 3 are set out in Table 3. The estimate of is positive and signi�cantly so,
suggesting an initial 11% monthly increase in sales for early adopters. The results for
the individual-speci�c �xed e¤ects, �si, are graphically depicted in Figure 1 for the 25
households for which we have sales data for kattumaram as well as �bre boat �shing.
Each of the 25 data points has abscissa equal to b�Ci and ordinate b�Fi. Notice that, forthose �shermen who adopted before t0 + 5, b 1Di(5) has been added to b�Fi. The diagramthus gives the long-run expected gains from technology adoption, which will also be used
throughout the rest of this paper. The straight line depicts the 45� line. According to
these results, three �shermen su¤ered a loss in sales of more than 1%, 2 experienced
virtually no change (less than 1% change), while 20 enjoyed increases in average sales
between 3.5 and 158%. The average change equals 40.2% with a standard deviation of
46.8%.
6.2 Determinants of the Timing of Technology Adoption
When a technology is divisible, like the adoption of new seeds in agriculture, a farmer with
several plots can choose on how many of them to try the new technology. In contrast,
a �shing boat is by nature an indivisible productive asset for a household. Moreover,
switching technologies is expensive, while with many technologies previousl studied in
agricultural contexts, a farmer can reverse the technology switch in subsequent growing
cycles without incurring a cost from switching back. To summarize, in the context of
5Notice that the statistical properties of the point estimate of � are non-standard as minimization of
the sum of squares over � is a discrete problem. Therefore Table 3 only contains the point estimate of �.6The estimte of � can be reconciled with the estimates of equation 2, which suggest that learning by
doing lasts for twice as long. Notice that the quadratic function used there is downward sloping for high
values of �i and thus leads to an upward biased estimate of the duration of learning if the learning curve
is in fact �at for high values of �i.
16
adoption of new crop varieties in agriculture, the adoption decision is typically both
divisible and reversible, while in the present setup, neither of these two properties holds.
Since adoption in the context of this study can be interpreted as a one-time transition
from one state, kattumaram �shing, to another state, �bre boat �shing, the timing of the
individual adoption decision is most suitably modelled using methods from the statistical
analysis of survival data. For the estimation, we adopt the common proportional hazard
assumption. According to it, the hazard �, that is the probability that i adopts within
the next period given that he has not adopted yet, can be factored into a baseline haz-
ard function, which is the same for all individuals in the population, and a function of
individual characteristics, xi. Speci�cally, it is assumed that
�i(t) = �0(t) exp(x0i�);
where � is a vector of parameters. From this structure of individual hazard, the likelihood
of each observed adoption time can be derived as a function of the adoption time t�i , xi and
�. An expression for the likelihood can be obtained regardless of whether or not adoption
occurred before the date of the interview. When the latter is true, the observation is
treated as �censored�. Using Cox�s (1975) semiparametric method of partial likelihood,
maximum likelihood estimates of � can be obtained numerically without making any
functional form assumptions about the shape of �0(t).
An individual with characteristics xi has a hazard higher than the sample average if
she is more likely to adopt earlier than the average of the sample because she faces a
higher probability of switching at any time t0 after date zero, conditional on not having
switched already before t0. The sign of the relationship between an explanatory variable,
xik say, and the outcome variable t�i thus goes the opposite way from an OLS model in
which adoption time is regressed on xi: in the proportional hazard model, a positive value
of �k implies that an individual with a higher value of xik faces a higher probability of
making the transition at any given point in time, and thus reduces the expected value of
his adoption time, t�i . In the OLS model, in contrast, a positive value of �k implies that
an individual with a higher value of xik adopts later in expectation.
From the model of the previous section, one key explanatory variable of interest is the
income gain that an individual expects from the transition. Recall that, in our simple
17
model, an individual starts saving to �nance the new technology as quickly as possible
only if the expected net gain from adoption is positive. Unfortunately, the researcher does
not observe individual expected net gain but only a measure of realized net gain, which
can be retrieved from b�Fi and b�Ci. We interpret realized net gain as a proxy for expectednet gain. More speci�cally, when individual expectations are unbiased, realized net gain
equals expected net gain plus a random error term which has expectation zero. De�ne
�yi = exp(�F )� exp(�C)
as the proxy for expected net gain in absolute terms. When �yi is included as a regressor
in the vector xi, however, we potentially face the problem of a contaminated regressor for
at least two reasons. First, the applicable explanatory variable is expected net gain while
the variable used is a noisy realization of it. We are thus facing a problem analogous to
the one of errors in variables in a linear regression model. The extent of the estimation
bias induced by this problem depends of course on how accurate individual expectations
are. If individuals can perfectly predict the actual income change, the use of �yi as
explanatory variable is valid. The wider realized gains are distributed around expected
gains, the more severe the bias introduced by using �yi.
Second, individual shocks may not be i.i.d. in each month, but rather be correlated.
For example, if a �sherman falls unexpectedly sick for an extended period of time right
after purchasing a �bre boat and this reduces his ability to go �shing, �yi underestimates
his expected gains.
For both of these reasons, we will experiment with two speci�cations in the empirical
analysis. One where �yi is included in the xi vector without modi�cation and one where,
in the spirit of a two stage least squares model, �yi is �rst regressed on a vector of
instruments and its predicted values, c�yi say, are used as explanatory variable in thesubsequent regression of the timing of adoption. Indirect evidence for the �noisiness�
of �yi is provided by the fact that our estimates of �yi are negative for one �fth of
those households for which both kattumaram and �bre sales data are available. For these
households, individual rationality seems to be violated as they adopt although they expect
smaller pro�ts from the new technology.
18
To illustrate how income change and adoption time are empirically related, Figure
2 plots t�i over �yi. If �yi is an accurate measure of expected gains, unconstrained
economic e¢ ciency dictates that all households which realize a positive income change
adopt immediately while those with a negative �yi never adopt. When funds available
to the �shing village are limited, constrained economic e¢ ciency dictates that households
which realize a positive income change adopt in decreasing order of �yi. While there is
some negative correlation between t�i and �yi (the correlation coe¢ cient equals -0.04),
this relationship is weak and statistically insigni�cant.
Another set of key explanatory variables refers to the capital market conditions a
household faces. Here we consider two categories, income and asset variables. Within
the �rst one, yCi = exp(�Ci), average sales generated with the old technology, proxies a
household�s income stream before adoption. If the technology switch requires own funds
that are not present when the new technology becomes available, a household with higher
yCi will be able to accumulate the required own funds faster. A signi�cant negative
relationship between yCi and t�i can thus be taken as evidence for a credit constraint faced
by an income-poor household. Another income variable that will be used is the number
of household members who earn an income.
The second one, the value of the house at the time when the new technology became
available, is an important component of the assets a household can collateralize to obtain
credit. A signi�cant negative relationship between a0i and t�i can thus be taken as evidence
for a credit constraint faced by an asset-poor household. Other variables that will be
initially included are household size as well as the household head�s literacy, age, both
linear and squared, and years as boatowner as a measure of experience.
Table 4 gives the results of the estimation of the determinants of adoption timing.7
Column 1 gives coe¢ cient estimates together with asymptotic p-values for the full set
of regressors, including �yi not instrumented.8 At conventional signi�cance levels, only
7Notice that Cox�s method of partial likelihood does not identify an intercept term.8For the three censored observations in the sample used for this estimation we have to impute values
of �yi. These are obtained by regressing �yi of the available 23 uncensored observations for which we
have both yCi and yFi = exp(�Fi) on house value, yCi, age, age squared, literacy and number of crew
members who belong to the extended family, and using the estimated coe¢ cients to generate predicted
19
the value of the �sherman�s house is a signi�cant determinant of the timing of �bre boat
adoption. The positive sign of the coe¢ cient means that a wealthier (in terms of assets)
household is more likely to adopt the new technology earlier. Of the two variables that
proxy for the income status of the household, yCi is signi�cant at the 12% level while
the number of family members who earn an income is insigni�cant. The same applies for
household size and age. AWald chi-square test of the hypothesis that both age coe¢ cients
are equal to zero fails to reject with a p-value of 0.58.
Column 2 gives coe¢ cient estimates for a speci�cation that uses predicted values of
�yi, c�yi, for the entire sample. As elaborated above, the concern addressed with thismethodology is that there are reasons to believe that �yi is a noisy realization of the
income change expected by an individual. The problem, however, is to �nd good instru-
ments for �yi that do not a¤ect the timing of adoption directly. The best one we could
�nd in our data is the number of crew members employed by the head of household who
belong to the extended family. It is, however, still a rather weak instrument. The only
two noticeable changes with this estimation procedure are, �rst, that yCi is now substan-
tially less signi�cant and, second, that our measure of experience, years as boatowner,
becomes more signi�cant. Finally, the Wald chi-square test of the hypothesis that both
age coe¢ cients are equal to zero fails to reject with a p-value of 0.92.
Guided by the �ndings of speci�cations 1 and 2 and in regard of the fact that the
sample underlying this estimation is small, we also estimate a more parsimoneous version
where the four least signi�cant explanatory variables are omitted. According to column 3
of Table 4, both asset and income poverty signi�cantly delay adoption. Households with a
greater realized income gain are likely to adopt earlier, but this relationship is signi�cant
only at a level of 0.16. As before, greater experience in kattumaram �shing induces earlier
adoption.
Column 4, where the income change is instrumented, con�rms these �ndings. As in the
full speci�cation, instrumenting mainly a¤ects the coe¢ cient on yCi, which ceases to be
signi�cant at conventional levels in this speci�cation. To summarize columns 1 through 4,
we �nd compelling evidence that asset poverty delays adoption and mixed evidence that
values of �yi:
20
income poverty does so as well. On the other hand, households that can expect a larger
income change from adoption are not more likely to adopt earlier.
6.3 The Role of Wealth
We now discuss in some detail how asset wealth a¤ects the timing of adoption. We
start by considering the arguments of Besley and Case (1994) and Foster and Rosenzweig
(1995) that asset wealth accelerates adoption because land-rich households enjoy higher
intertemporal bene�ts from experimentation due to their larger scale of operation. In
our sample, in contrast, each household operates exactly one boat before and after the
switching of technologies, so that we can safely discard the scale argument.
Another channel we can con�dently rule out is that wealthy households adopt earlier
because of better access to the labor market. In the setup studied here, the same amount
of labor is employed to operate the old and the new technology. Each household in our
sample which adopts the new technology has operated the old technology before and thus
already secured the amount of labor needed for the new technology.
What about better access of wealthier households to the new technology? Each house-
hold in the sample obtained its FRP from the nearby branch of a domestic FRP manufac-
turer. That branch is less than 4 kilometers away from the village and no transaction costs
for transportation are incurred from the purchase. Moreover, according to villagers, there
has never been a supply constraint ever since the new technology has become available in
2000. It can thus be ruled that wealth works through overcoming a supply constraint or
having enhanced access to the new technology.
We next examine the relationship between initial wealth and risk-bearing attitudes. It
is commonly believed that preferences for risk bearing crucially depend on a household�s
wealth. In particular, under the plausible assumption of decreasing absolute risk aversion
(DARA), households above a certain wealth level choose to incur a given lottery with
positive expected payo¤while households with wealth below that level choose to stay away
from it, although they would accumulate assets to later choose the lottery. Apparently,
adoption of an FRP entails two forms of risk.
First, the amount of �sh catches �uctuates from day to day depending on weather and
21
maritime conditions as well as individual luck. The question, however, is whether these
�uctuations are more severe with an FRP than with a kattumaram. To obtain an answer,
we run the regression
log(ysit) = �si + �t + usit
separately for s = C and s = F . The resulting root mean squared errors are 0.66
and 0.50, respectively. Thus, controlling for scale by considering the natural logarithm
of sales, operating an FRP entails a smaller month-to-month risk than a kattumaram.
While it may be argued that daily catches may exhibit di¤erent volatility patterns across
technologies than monthly ones, it is not likely that those are particularly relevant as
informal insurance arrangements seem to be prevalent in these villages. In this connection,
boatowners report that they can easily obtain a short-term consumption loan from their
auctioneer to compensate for a series of bad catches.
Second, as pointed out in the previous subsection, a kattumaram operating boatowner
may face uncertainty about the level of average gains (net of day-to-day �uctuations)
from the technology shift. This together with the DARA assumption can explain later
adoption by poorer but ex-post equally successful households. This explanation competes
with the remaining one of credit constraints. Since our quantitative data cannot provide
a de�nite answer in favor of either one of the two, we will use additional, perceptional
data to get a sense of the relative importance of each of the two competing hypotheses.
Our survey asked each boatowner the following question: �Why did you wait (are you
waiting) to switch to a FRP boat? Give the most important reason.�. By far, the two most
frequent answers were, �rst, �It required a lot of capital�, and second, �I was uncertain
about the bene�ts�. Table 5 gives some statistics relating to the characteristics of the
respondents by their answer to this question. The pattern we �nd is as follows. First,
the capital requirement is mentioned roughly 50% more often than bene�t uncertainty.
Second, wealth among those who cite bene�t uncertainty as the main reason is on average
more than 25% higher than among those who mention the capital requirement �rst. This
suggests that the capital constraint is more severe for poorer entrepreneurs, in fact to
such an extent that it dominates the concern about bene�t uncertainty, even though that
latter concern is also of greater importance to poorer decision makers when DARA is
22
postulated. While the di¤erence in asset wealth across answers is on the order of 30%,
this di¤erence is not statistically signi�cant. In that light, we do not have statistically
signi�cant, albeit economically important, evidence for the assertion that a lack of wealth
a¤ects the timing of adoption mainly through limited access to capital.
7 Simulation
The �ndings of the estimation suggest that asset poverty delays technology adoption. To
be more precise, among two households which expect the same increase in average income
from adoption, the wealthier one is more likely to adopt �rst. In this section, we address
the policy-relevant question of how alternative distributions of wealth, as measured by
house value, change the pattern of technology di¤usion. We focus on the relationship
between the wealth distribution, which will be a¤ected by the di¤erent economic policies
considered, and the outcome variables mean income (within the sample) and income
inequality.
To conduct simulations, we �rst need to specify a baseline hazard function, �0(t). We
make the assumption of a constant baseline hazard,
�0(t) � �;
given the small sample we have. Moreover, we consider a situation in which each household
adopts exactly at the expected value of its adoption time,
bt�i = E[t�i jxi];which is of course a function of b�. With a constant baseline hazard, we obtain
bt�i = e�x0i b�=�:Finally, the parameter � is calibrated as follows. In our sample, three households have
not adopted before the interview date. We thus choose � such that the date of the last
adoption recorded before the date of the interview matches the fourth to last adoption
date in the data simulated with the actual values of xi.
23
Figure 3 plots actual and simulated mean income. Notice that actual mean income
uses all ysi for �xed t, that is yFi (yCi) enters the average when household i has (not)
adopted before date t. More formally, actual mean income is computed as
1
n
nXi=1
({ft < t�i gyCi + {ft � t�i gyFi) :
The formula for predicted mean income is given by the same expression, except that t�iis replaced by bt�i . The predicted data is generated from the speci�cation of Column 3
in Table 4. Without reproducing the results separately, we note that the shape of the
predicted graph remains qualitatively unchanged when the instrumented version, Column
4 in Table 4, is used instead.
According to the solid line in Figure 3, there are three obvious �waves�of adoption:
at the beginning, then just before one year later, and �nally a little more than two years
later. Notice that the solid line ends at the 36th month, the last date for which we have
data. Our simulation model appears to capture satisfactorily the main features of the
data, though the predicted path is smoother than the stair-shaped pattern in the actual
data. According to the simulation, the last household in the sample adopts 54 months
after the technology has become available. At that time, predicted average income has
increased by about 39%.
Figure 4 depicts the Gini index of estimated actual incomes and the Gini as predicted
by the simulation model. Notice that inequality during the adoption process exhibits
the familiar inverted U shape. This re�ects, �rst, that on average adopters experience
a substantial increase in income and, second, that it is not the initially income-poor
who adopt �rst because in that case adoption would narrow the income gap between the
initially income-rich and poor. In the data, we see an increase of the Gini from 0.34 to
0.38 during the �rst wave of adoptions. The second wave of adoptions a year later leaves
inequality virtually unchanged, while the third wave results in a drop of the Gini of about
20% to a level of 0.31, which is substantially lower than the value that prevailed before
the new technology was known. All in all, while the village experiences a substantial
increase in inequality over a course of two years, the availability of the new technology
can hardly be criticized for its long-term impact on the village economy since, at the same
24
time, average income increases and inequality decreases substantially.
The predicted data satisfactorily captures the main features of the data. It correctly
predicts the jump in inequality induced by the �rst wave of adoptions. The consequences of
the second and third wave, however, are less clearly distinguishable in the simulated data,
because, according to the dotted line, inequality gradually decreases from the eleventh
month onward. The last predicted adoption in the 54th month leads the village to a Gini
of 0.285, which is sixteen percent lower than the one at date zero, where all households
operate the old technology.
We now turn to the simulated policy counterfactuals. We �rst investigate the con-
sequences of redistributive policies. Toward this, we assume that each household in the
sample holds just the mean level of wealth observed in the data, i.e. owns a house worth
Rs. 75,380. In such a scenario, the credit constraint is loosened for households whose
wealth is below average and tightened for the rest. If the relationship between wealth that
can be collateralized and the extent to which a household is credit-constrained is concave,
we expect adoption to occur more promptly on average with such a policy in place. The
results for mean income and the Gini are plotted in Figures 5 and 6, respectively. Ac-
cording to Figure 5, equal redistribution does in fact result in a quicker adoption process.
According to the simulation, the last adoption occurs a year earlier, in the 42nd instead
of the 54th month, than with the actual wealth distribution. The e¤ect on sales over the
course of the adoption process, on the other hand, is rather small. With an equal asset
distribution, simulated sales never exceed predicted actual ones by more than 7 percent.
Moreover, when we focus on di¤erences between simulated and predicted actual sales of
more than 3%, simulated sales never lead predicted actual ones by more than �ve months.
According to Figure 6, a similar picture emerges for the dynamics of inequality. While
the inverted U contracts by about 20% toward the origin, the change in the general pattern
of inequality as measured by the Gini can hardly be judged economically signi�cant.
A second set of simulations investigates two extreme scenarios. The �rst one assumes
that each household in the sample holds only the smallest observed wealth, that is each
house is assumed to be worth Rs. 20,000. The second one, in contrast, assumes that
each household in the sample holds the highest observed wealth, that is each house is
25
assumed to be worth Rs. 500,000. The results for this set of simulations together with
the predicted actual values are set out in Figures 7 and 8. We thus consider situations in
which all households are either tightly credit-constrained or virtually do not face a credit
constraint at all. The mean income and inequality paths for the �rst simulation very
closely follow the respective paths generated from the actual asset data, which suggests
that the observed income pattern accompanying the introduction of the new technology
closely resembles a situation in which all households are substantially credit-constrained.
The results for the second simulation, where the credit constraint is released for the
entire sample, are more striking. The dotted lines in Figures 7 and 8 suggest that with a
uniformly high level of asset wealth the adoption process is completed in just �ve months.
As a consequence, the village enjoys a substantially higher mean income for about two
years by which adoptions in the simulated data lead predicted actual ones. This result
suggests that a community in which households face virtually no credit constraints is able
to move up the technology ladder much faster than the one investigated by this study.
Similarly, only a minor spike remains of the observed pronounced inverted U shape of
inequality.
8 Conclusions
This paper studies the di¤usion of a new technology among south Indian �shermen, which
is as much a product of ongoing globalizing trends as it is a response to distortions caused
by previous waves of innovation triggered by globalization. We identify determinants of
the timing of technology adoption as well as resulting income and inequality dynamics
during this process. We �nd that lack of wealth is a key predictor for delayed adoption and
that the channel through which this mechanism is e¤ective is a credit constraint. During
the di¤usion process, inequality follows Kuznets�well-known inverted U-shaped curve.
Simulations suggest that a redistributive policy favoring the poor results in accelerated
economic growth and a shorter duration of sharpened inequality, although the quantitative
impact of such a policy is small.
One advantage of this paper over other studies is that context is well understood. Thus,
26
the speci�c channels in which wealth matters for adoption, credit constraints as well as
higher risk aversion, are identi�ed. We conclude, like Platteau (1984), that overall our
study village experienced a success story of globalization. According to our simulations,
technology di¤usion for the entire sample is completed in less than �ve years and income
gains for the initially poor are relatively larger than for the rich.
What remains unaddressed by this research are the long-run consequences for the
resource base and thus future generations of �shermen due to increased e¢ ciency in
�shing. Future work will have to evaluate whether the short-term gains generated by
the di¤usion of �bre reinforced plastic boats are both economically and environmentally
sustainable. Previous instances of globalization and subsequent resource depletion in low
income countries warrant scepticism.
References
[1] Bandiera, O. and I. Rasul, 2004. Social Networks and Technology Adoption in
Northern Mozambique, unpublished manuscript, London School of Economics.
[2] Besley, T. and A. Case, 1994. Di¤usion as a Learning Process: Evidence from HYV
Cotton, Princeton University Department of Economics Working Paper 5/94.
[3] Binswanger, H., J. Dayantha, T. Balaranaia and D. Sillers, 1980. The Impacts of
Risk Aversion on Agricultural Decisions in Semi-Arid India, working paper, World
Bank, Washington DC.
[4] Cox, D. R., 1975. Partial Likelihood, Biometrika 62, 269 - 276.
[5] Feder, G., R. E. Just and D. Zilberman, 1985. Adoption of agricultural innovations
in developing countries: a survey, Economic Development and Cultural Change 33,
255-297.
[6] Feder, G. and G. T. O�Mara, 1982. On information and innovation di¤usion: a
Bayesian approach, American Journal of Agricultural Economics 64, 145-147.
27
[7] Foster, A. and M. Rosenzweig, 1995. Learning by doing and learning from others,
Journal of Political Economy 103, 1176-1209.
[8] Giné, X. and S. Klonner, 2005. Dynamic Lending with Limited Commitment and
Uncertain Borrower Types: Theory and Evidence, working paper, Cornell University
Department of Economics.
[9] The Hindu, 2001. Boon for small-scale �shermen, The Hindu, March 22, 2001.
[10] Kurien, J., 1994. Kerala�s marine �sheries development experience. In B. A. Prakash,
editor, Kerala�s economy: Performance, problems, prospects, 195-214. New Delhi:
Sage.
[11] Kurien, J., 1995. Plywood boats in South India, Appropriate Technology 22, (2),
22-49.
[12] Munshi, K., 2003. Social Learning in a Heterogeneous Population. Technology
Di¤usion in the Indian Green Revolution, unpublished manuscript, Brown University.
[13] Nissanke, M. and E. Thorbecke, 2005. Channels and Policy Debate in the
[14] Platteau, J.P., 1984. The Drive Towards Mechanization of a Small-Scale Fisheries
in Kerala: A Study of the Transformation Process of Traditional Village Societies,
Development and Change 15, 65-103.
[15] Pietersz, V. L. C., 1993. Developing and Introducing a Beachlanding Craft on the
east coast of India, Bay of Bengal Programme, Madras, India.
[16] Sandven, P., 1959. The Indo-Norwegian Project in Kerala, Norwegian Foundation
for Assistance to Underdeveloped Countries, Oslo.
[17] Vivekanandan, V., 2002. The Introduction And Spread Of Plywood Boats On The
Lower South-West Coast Of India, Technical Report, South Indian Federation Of
Fishermen Societies.
28
29
Table 1. Descriptive statistics for the core sample
Mean Std Dev. Minimum Maximum
Sample Size
Value of House (in thousand Rs.)
Number of Family Members with other Income Source
Average Monthly Fish Sales before Adoption (Rs.)
Change in Monthly Sales from Adoption (Rs.)**
Household Size
Literacy of Household Head*
Age of Household Head
Years as Boatowner
Adopted FRP before January 2004
Adoption month**
26
75.38
2.00
22052.45
8419.69
6.42
0.38
38.46
10.57
0.88
Jan. 2002
97.74
1.01
15860.84
10550.01
3.03
0.49
12.12
5.06
0.31
8.88
20.00
1.00
5497.34
-8750.07
3.00
0
21.00
3.00
0
Jan. 2001
500.00
5.00
76017.63
48339.28
17.00
1.00
65.00
20.00
1
March 2003
* equals one if he reports that he can read or write, and zero otherwise. ** for those households that had adopted before the interview, which took place in the 62nd month. Table 2. Estimation results for equation 2
R-Square 0.694 No. of obs. 1471 Notes: Coefficients for 60 monthly dummies and 30 individual-specific fixed effects for kattumaram-operating fishermen as well as 42 individual-specific fixed effects for fibre boat-operating fishermen not reproduced
30
Table 3. Estimation results for eq. 3
Parameter Estimate
StandardError T p
κ 5
Di(5) 0.111 0.039 2.89 0.004
R-Square 0.692 No. of obs. 1471 Notes: Coefficients for 60 monthly dummies and 30 individual-specific fixed effects for kattumaram-operating fishermen as well as 42 individual-specific fixed effects for fibre boat-operating fishermen not reproduced Table 4. Determinants of the timing of adoption. Dependent variable: month of adoption
(1)* (2) (3) (4) Value of House 0.00525
(0.070)0.10697(0.064)
0.00557 (0.022)
0.00528(0.026)
Family members with Income
-0.17937(0.777)
0.00521 (0.860)
Average Income before Adoption
0.0000429(0.122)
0.06286(0.439)
0.0000414 (0.057)
0.0000287(0.158)
Income Change from Adoption
0.0000506(0.151)
0.0000187(0.786)
0.0000346 (0.161)
-0.0000074(0.912)
Household Size 0.03444(0.868)
-0.0000289(0.760)
Literacy of Household Head
-0.86901(0.156)
-0.72087(0.246)
-0.73734 (0.173)
-0.79233(0.138)
Age of Household Head
-0.22363(0.311)
-0.02748(0.919)
Age Squared 0.00273(0.298)
0.0004796(0.878)
Years as Boatowner 0.10930(0.222)
0.12437(0.149)
0.11890 (0.105)
0.14000(0.065)
Log-Likelihood -47.9 -48.9 -48.5 -49.3Income Change Instrumented No Yes** No Yes**
Number of Obs. 26 26 26 26No. of Obs. Censored 3 3 3 3
* Asymptotic p-value in parentheses ** Instruments: Age, age squared, years as boatowner, number of crew members who belong to the extended family
31
Table 5. Wealth status by self-reported reason for delay of adoption, core sample Answer N Mean Std Dev Minimum Maximum
Capital Requirement
13 69.2 63.7 0 250
Benefit Uncertainty
9 95.5 152.7 20 500
Other 4 Figure 1. Individual average profitability with fibre boat over individual average profitability with kattumaram for 25 households for which sales data is available for both kattumaram and fibre boat fishing
MU1
8
9
10
11
12
MU0
7 8 9 10 11 12
32
Figure 2. Adoption date over realized absolute income change for 25 households for which sales data is available for both kattumaram and fibre boat fishing
corr_dat e_f b1_num
20
30
40
50
60
ydi f
-10000 0 10000 20000 30000 40000 50000
Figure 3. Mean income after the new technology became available, actual (dotted line) and predicted by the model (dashed line)
meanreal
22000
23000
24000
25000
26000
27000
28000
29000
30000
31000
MONTH
0 10 20 30 40 50 60
33
Figure 4. Income Gini after the new technology became available, actual (dotted line) and predicted by the model (dashed line)
gi ni real
0. 28
0. 29
0. 30
0. 31
0. 32
0. 33
0. 34
0. 35
0. 36
0. 37
0. 38
0. 39
0. 40
MONTH
0 10 20 30 40 50 60
Figure 5. Predicted actual (solid) and simulated (dotted) mean income. Simulation assumes perfectly equal distribution of wealth (measured by house value) over the sample
meanpred
22000
23000
24000
25000
26000
27000
28000
29000
30000
31000
MONTH
0 10 20 30 40 50 60
34
Figure 6. Predicted actual (solid) and simulated (dotted) Gini. Simulation assumes perfectly equal distribution of wealth (measured by house value) over the sample
gi ni pred
0. 28
0. 29
0. 30
0. 31
0. 32
0. 33
0. 34
0. 35
0. 36
0. 37
0. 38
0. 39
0. 40
MONTH
0 10 20 30 40 50 60
Figure 7. Predicted actual (solid) and simulated mean income. Simulation 1 (dashed) assumes the lowest observed wealth (house value equal to 20) for the entire sample, simulation 2 (dotted) assumes the highest observed wealth (house value equal to 500) for the entire sample
meanpred
22000
23000
24000
25000
26000
27000
28000
29000
30000
31000
MONTH
0 10 20 30 40 50 60
35
Figure 8. Predicted actual (solid) and simulated income Gini. Simulation 1 (dashed) assumes the lowest observed wealth (house value equal to 20) for the entire sample, simulation 2 (dotted) assumes the highest observed wealth (house value equal to 500) for the entire sample
